© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 10-RES_POST_32K-OAI
    11. 30010
    Prev
    Next
    INDEX
    Explanations

    references to the size or scale of entities or phenomena

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_10.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.10.hook_resid_post
    Hook Layer
    10
    Architecture
    standard
    Context Size
    64
    Dataset
    Skylion007/openwebtext
    Activation Function
    topk
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    68
    +0.10
    0.5%
    200
    +0.10
    0.4%
    586
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    623
    +0.10
    0.05
    67
    +0.10
    0.05
    669
    +0.09
    0.05
    Negative Logits
    ISON
    -0.65
     Exile
    -0.58
    Ĥ¬
    -0.58
     Directory
    -0.58
     embodiment
    -0.57
    ISSION
    -0.57
    KA
    -0.57
     Canaver
    -0.56
     Inquiry
    -0.56
     Malt
    -0.56
    POSITIVE LOGITS
     quantities
    1.80
     amounts
    1.77
     sums
    1.59
     percentages
    1.49
     numbers
    1.42
     portions
    1.40
     volumes
    1.37
     chunks
    1.37
     tracts
    1.27
     distances
    1.21
    Activations Density 0.832%

    No Known Activations